Semantic object recognition for BIM reverse engineering

Document Type

Conference Proceeding

Publication Date

1-1-2012

Abstract

Building Information Modeling (BIM) is becoming widely used in the construction and building management industries. BIM building models provide object-oriented 3D modeling of building components, as well as the ability to attach additional data and documents to building components. Creating BIM models for existing buildings, and maintaining the models throughout building lifecycles, is a difficult and time-consuming task. In this research, we propose an object identification framework for image and video which will identify objects and wall configurations and their special relationships in commercial office spaces for comparison to and updating of BIM models of the spaces. In particular, we will investigate the use of contextual and common sense knowledge in improving object recognition and segmentation process in such environments using deformable model and shape grammar.

Publication Source (Journal or Book title)

62nd IIE Annual Conference and Expo 2012

First Page

1933

Last Page

1938

This document is currently not available here.

Share

COinS